How does Cloudimage accelerate images?

There are 2 factors that make the loading times long :

• the size of the file,
• the network quality between the Origin Server and the End-User

By adapting the size/weight and format of your images and using his globally distributed multiple CDN networks, Cloudimage will improve the loading time of your images all around the world.

Process for the first time an image is requested using Cloudimage:

1. This image does not exist in the CDN as an optimized version
2. It requires a long process of:
   - downloading the Original image to our servers
   - transcoding the image (if GIF) - this means breaking the image to image per image - optimizing what is possible and compiling again (based on selected size and parameters/functions
   - the new version to be delivered to CDN
   - from there the end-user to get it
3. Once the image version (resized copy) is cached in our CDN it gets directly delivered to your end-users - skipping all other steps
4. The best way to compare the speed difference is:
   - Open a new image for first time in your browser while observing the network on the browser dev tab
   - Do hard refresh (reload with cleared cache)
   - Compare the time between the 2 requests

The bigger Origin file and the more complex transformations needed == the longer TTFB for the first delivery

When an image is requested we have 3 possible cases:

1. The image is requested for first time, we do not have the transformation in our CDN and don't have the Original file 
2. The CDN have not cached the requested image resize, but the Cloudimage servers have previously downloaded the Original file
3. The CDN have available the requested image resize.

As visible here we have all 3 points listed in the example image - we can see the Time for each

And here we can see in detail what process took time for each of the 3 requests:

• First time requesting file - (all listed process above)
• Requesting new resize (already available Original file)
• Requesting cached resize 

If you have any questions or issues, please feel free to contact our Software Engineer and mention the following details:

• Token
• Short description of the issue (including URLs, screenshot, short video if available)
• Steps to reproduce the issue.